By Katy Cain
When it comes to earthquakes, seismologists know best. But deep-focus earthquakes—those that occur between 300 and 700 km below the surface—have been a mystery to science for a century. Over the years, many scientists have suggested different answers to this problem. Still, the deep Earth is impossible to study directly, so the earthquakes kept their secrets until a group of scientists at the Earth and Planets Laboratory teamed up to crack the case.
The majority of earthquakes occur close to the Earth’s surface, down to about 70 km. The shaking starts when stress builds up at a fracture between two blocks of rock—known as a fault. As you go deeper into the planet, the intense pressures in the deep Earth create such high friction along faults that it becomes hard to explain how rocks could slide past each other in the same way.
In fact, scientists assumed that it was physically impossible for earthquakes to occur past the 70 km mark until the 1920s when Japanese seismologist Kiyoo Wadati found seismological proof of quakes occurring much deeper. These deep quakes are found so exclusively at or near subduction zones that they used to be their defining feature. Subduction zones occur where one tectonic plate—normally hydrated oceanic plate—is dragged beneath another as it sinks deep into our planet’s mantle.
More than a century later, the cause of the planet’s deepest earthquakes is still a mystery to science.
There are two tiers of these deep earthquakes. Intermediate-depth earthquakes occur in the mantle between 70 and 300km below the surface. Scientists are fairly certain that dehydration of minerals as they descend into the planet is a major source of these quakes. On the other hand, deep-focus earthquakes occur between 300 and 700 km below the surface. The source of these earthquakes is still a hot topic of debate among Earth scientists. Could a similar dehydration mechanism explain the deep-focus quakes? As people studied the problem, models seemed to agree that water-bearing minerals would be “wrung dry” by the time the slabs descended past 300 km.
Said Carnegie seismologist Lara Wagner, “People moved away from thinking water played a role in deep-focus earthquakes because they didn't think water-holding minerals could make it that far down.”
Where there are diamonds there’s water
In 2017, geochemist Steve Shirey and seismologist Lara Wagner were in attendance at a particularly rousing seminar at the Earth and Planets Laboratory. The group was discussing a 2007 paper by high-pressure mineralogist Harry W. Green II that investigated the source of the deep-focus earthquakes.
In that paper, Green suggested that earthquakes past 300km would occur due to sudden high-pressure changes in olivine—a mineral that makes up the majority of the Earth’s upper mantle. Green (and many others) based this argument in part on the idea that water could not make it down to the depths of deep-focus earthquakes.
This seemed odd to Shirey because the work he was doing with “super-deep diamonds—diamonds that form up to 700 km—showed the exact opposite.
“I realized that the seismologists in the group didn’t know about our recent work or even any of the general work in the community on super-deep diamonds. So, I raised my hand,” said Shirey. “I remember telling them, ‘This can’t be right. We have diamonds from down there and that means they have to be forming in fluids.”
From that moment, both Shirey and Wagner knew they were onto something. They met up after the seminar and put their heads and datasets together, comparing the depths of rare deep-Earth diamonds to the mysterious deep-focus quakes. To their excitement, they realized that they had hit on a new observation: diamond-forming fluids and deep earthquakes in deeply subducted slabs were occurring in about the same place! Studying the inclusions trapped inside of the diamonds clearly showed two things 1) the diamonds were indeed forming between 300-700 km beneath the surface of the planet, and 2) the diamonds were forming from materials that traveled with the subducting slab from the surface all the way down to the deep Earth.
Gathering the dream team
In order to solidify the finding, Wagner and Shirey needed to sort out what conditions could allow for water and other fluids to travel to these inconceivable depths. For this task, they enlisted colleagues Peter van Keken, a geodynamicist who could model the subducting slabs, Michael Walter, an experimental petrologist who could sort out the chemistry behind what was happening to the rocks under such extreme pressures and temperatures, and Graham Pearson a former Carnegie postdoctoral fellow now from the University of Alberta who brought additional experience and expertise in deep diamond petrology. In fact, it is Pearson’s direct analyses of water in the super-deep diamond inclusion mineral ringwoodite that provided hard analytical evidence of deep water.
With this powerhouse of planetary science assembled, they were ready to tackle the question from all sides.
Said Shirey, “The nature of deep earthquakes is one of the top mysteries in science. It’s not often you get to work on something like that with your scientific friends.”
In textbooks, subducting slabs look like smooth sheets of rock sinking steadily into the Earth. But in reality, some slabs look more like torn sheets rippling in a strong wind. This poses a problem for models of slab temperatures as a function of depth because they depend on the assumption that the slabs are smooth and continuous.
To ensure that the study used only models that were true to their underlying assumptions, Wagner began with a compilation of cross-sections at even intervals across all subduction zones, and then removed those profiles that were too close to changes in slab shape or slab tears. This was vital to the study because, with the right selections, the team was able to see through the noise caused by complex subduction zones to see a clear pattern between the occurrence of deep earthquakes in some slabs and the temperatures of those slabs.
Said Wagner, “We limited ourselves to very simple subduction geometries that we could accurately model the temperatures in two dimensions.”
Wagner’s carefully chosen 23 subduction zones were used by van Keken to develop advanced computational models of the slabs’ temperatures at much greater depths than any previous studies of this nature.
The next step was to take these slab temperatures and pressures and to compare them to the pressures and temperatures at which water-bearing minerals are stable. Walter looked at a range of different model parameters and different depths within the downgoing plate and compared these to phase diagrams of the types of rocks known to exist in those parts of the subducting slab. He found clear evidence that those slabs that did have deep earthquakes were able to transport water to the depths of these events, whereas slabs that did not have deep earthquakes lost all of their water at much shallower depths.
“We assessed the temperature profiles for all of these different places in the world and we compared them to the stability of these different carbon- and water-bearing mineral phases to see if the slab could bring water and carbon to those depths,” said Wagner. “What we found is that some slabs could, and some slabs couldn’t.”
The work, which was recently published in AGU Advances, provides the first robust evidence that fluids play a key role in deep-focus earthquakes.
Only at Carnegie Science
Just as exciting as the finding itself was how smoothly each new piece of evidence fit into place. Each new piece of the puzzle fit perfectly—a rarity.
Wagner mused, “When you do big survey studies like this, you expect it to be messy. Some slabs work, some don't, maybe you have to statistically prove that what you found is relevant,” Wagner continued, “What really struck us was that the first time we plotted up these results, we looked at them and were shocked by how clear it was. That doesn’t happen very often.”
According to Shirey, collaborative papers like this one are what give Carnegie Science its raison d’etre. Carnegie Science actively encourages its scientists to seek out opportunities to work together across disciplinary boundaries to tackle new problems where they can. Scientists at Carnegie are also free to pursue scientific questions as they come up, with the support of the institution they don’t have to wait on grant money to get started.
In 2020, Carnegie’s historic Department of Terrestrial Magnetism and Geophysical Laboratory merged to create the Earth and Plants Laboratory. The merge solidifies the two departments as one team, and this paper is just one example of the top-tier work that can come from cross-campus collaboration.
Said Shirey, “We needed all four of these different disciplines to come together to make this argument. It turned out we had them all in-house! It's rare to find a department that has that breadth.”
- AGU Viewpoint by Douglas A. Weins: "Diamonds hold clues about deep focus earthquakes"
- AGU Editors Highlight: "Diamonds are at fault"
- Carnegie Science Press Release: "What causes Earth's most mysterious earthquakes?"
- Full Article in AGU Advances: "Slab Transport of Fluids to Deep Focus Earthquake Depths—Thermal Modeling Constraints and Evidence from Diamonds"
- Carnegie Science Press Release: "Deep Diamonds Contain Evidence of Deep Earth Recycling Processes"
- Carnegie Blog featuring co-authors Michael Walter and Graham Pearson: "Your diamond questions answered!"
- Related Article in Nature Geoscience: "Dry metastable olivine and slab deformation in a wet subducting slab"